针对带有线性等式和不等式约束的无确定函数形式的约束优化问题,提出一种利用梯度投影法与遗传算法、同时扰动随机逼近等随机算法相结合的优化方法.该方法利用遗传算法进行全局搜索,利用同时扰动随机逼近算法进行局部搜索,算法在每次进化时根据线性约束计算父个体处的梯度投影方向,以产生新个体,从而能够严格保证新个体满足全部约束条件.将上述约束优化算法应用于典型约束优化问题,其仿真结果表明了所提出算法的可行性和收敛性. In order to solve the constrained optimization problem with uncertain function in the form of linear equations and inequality constraints, this paper proposes an optimization method which uses a combination of gradient projection method and genetic algorithm, stochastic approximation and random stochastic approximation. This method uses genetic algorithm Search and utilize the simultaneous perturbation stochastic approximation algorithm for local search. The algorithm calculates the gradient projection direction of the father at each evolutionary time to generate a new individual, which can strictly ensure that the new individual satisfies all the constraints. The algorithm is applied to the classical constrained optimization problem. The simulation results show the feasibility and convergence of the proposed algorithm.